verma.unionimpactonhrm.jlr2005

8/8/2019 Verma.UnionImpactOnHRM.JLR2005

1/48

1

Forthcoming in the Journal of Labor Research, 2005.

What Do Unions Do to the Workplace?Union Impact on Management and HRM Policies

ANIL VERMA * University of Toronto, Toronto, Ontario, Canada M5S 3E6

I. Introduction

The objective of this paper is to build further on our understanding of union effects by examining

what unions do to managerial practice in the workplace. Unions can be an instrument of social

change but even when they play a larger role in society, their core activity remains focused at the

workplace. Their principal engagement is with management though their actions may extend to

lobbying, politics, and the community at both local and international levels. Therefore, in any

consideration of the question, what do unions do to the workplace, it is important to examine the

impact of unions on management in general and on human resource management (HRM), in

particular. The main focus for Freeman and Medoff, in their 1984 book, What Do Unions Do ,

(hereafter F-M) was not on this question but rather on union effects on outcomes such as

productivity. Their findings have been influential in advancing our knowledge of union impact on

organizational outcomes. They offer a number of explanations for their finding of a positive union

effect on productivity. Apart from lower quit rates, three other possible explanations are suggested:

seniority-based rewards, better job production standards (and better management accountability in

general), and more employer-employee communications (pp. 14-15). The latter two of these

explanations concern managerial practice but were not directly investigated in the study. This paper

fills a gap in research by examining empirical support (or lack thereof) accumulated since the early

1980s, for some of these and other explanations for a positive union effect on management practice

at the workplace level.


2/48

2

Citing a landmark study by Slichter, Healy and Livernash (1960) (hereafter, SHL), F-M

argue that unions can improve efficiency by putting pressure on management to tighten job-

production standards and accountability in order to preserve profits in the face of higher wages

(p.15). The SHL study put the issue of union impact on management practice in focus by examining

an exhaustive range of management policies in the workplace such as hours of work, wage

incentives, subcontracting, promotions and discipline, to name only a few of them . SHL found that

unions can have both positive and negative effects on management and hence on efficiency. Among

the positive effects of unions, SHL point to better management, a better balance between employer

and employee interests, and better communications. Although SHL did not formally define the term

themselves, their findings of a positive union effect has been referred to as the shock effect of

unions on management i.

Even though the SHL study reported both positive and negative union effects on managerial

practice, it is most often cited, only for unions positive effect on efficiency. Although some

efficiencies result from the union exercising its voice role, the SHL study cites numerous other

examples to show that union voice does not always contribute positively to productivity. In many

instances, union actions and policies have the effect of restraining productivity or efficiency.

Similarly, it can be argued that the monopoly face of unions can also contribute to restraining as well

as enhancing efficiency. Although the monopoly face is generally associated with negative outcomes

for management, some monopolistic policies of the unions can also generate efficiencies for

management. Thus, the relationship of union policies to efficiency, whether it be the voice or

monopoly face, is a complex one.

To further develop our understanding of union effects, it is important to examine empirical

evidence accumulated since the publication of F-M on union effects on management practice.

Studies conducted since the early 1980s allow us to examine union effects on a range of workplace


3/48

3

and HRM policies. Accordingly, the purpose of this paper is to re-visit the impact of unions on

managerial practice in HRM and related workplace practices. More specifically, this paper attempts

to fill two gaps in the research literature. First, it revisits the theoretical bases for union impact on

management. This section attempts to integrate and clarify a variety of explanations offered in the

past. Second, it updates the SHL study by examining the empirical evidence accumulated by various

studies since the early 1980s. The second half of the paper builds on the conceptual framework to

survey the empirical evidence accumulated on union effects on managerial practice in HRM and

other related workplace policies. In examining union impact, the focus of this paper is on managerial

practice rather than on employee or firm-level outcomes. For example, this paper does not include

topics such as the effect of unions on productivity or the effects of seniority on employee outcomes.

II. Conceptual Framework

Union impact on management is framed by two dynamic processes: formation of underlying

preferences of each party and the interaction between the two parties as each tries to pursue its goals.

When unions arrive on the scene by organizing workers, they tend to drive up wages. In the popular

mind, union preferences have been understood foremost in terms of obtaining more for their

members. Empirical evidence confirms that unions use their monopoly power to force employers to

pay better wages and benefits (Freeman and Medoff, 1984). However, unions do strive for and

achieve other goals that are equally important to them and their members, namely, fair treatment

from management. This aspect of unionism is especially important in considering the impact of

unions on management behavior. Demand for fairness leads the union to get into almost every area

of day-to-day managerial decision-making at the workplace level. Slichter (1941) called the

emergence of labour-management relations in the 1930s a system industrial jurisprudence in which

the union gains a voice in almost all aspects of management decision-making in the workplace that


4/48

4

have consequences for workers.

One possible response of management to an increase in wages is to substitute capital for the

more expensive labor. Even if the firm operates at the same level of overall productivity, i.e., a

combination of higher labor productivity and lower capital productivity, the ensuing outcome can be

sub-optimal because it would lead to inefficient allocation of resources. This can be called the price

effect of unions , i.e., the effect of an increase in the price of labor.

The price effect explanation assumes that the workplace was already producing at the

efficient frontier before unionization and that no further gains in productivity could be made ex-post .

A second set of explanations have been offered by a number of scholars who have argued that

management of its own does not generally operate at maximum efficiency. Notable among economic

theorists is Leibenstein who suggested the concept of X-inefficiency as an explanation for why

management would operate sub-optimally. This is discussed later in this section. Among industrial

relations researchers, the SHL study is widely considered to be the most comprehensive to gather

firm-level evidence to document an increase in managerial efficiency in the post-unionization period.

If we assume that management was operating sub-optimally prior to unionization, then it is

possible for management to be shocked into adopting policies that would extract new efficiencies

from the operations. This is the crux of the shock effect argument and the SHL study offers a

substantial body of detailed workplace-level evidence in its support. Although the SHL study is cited

frequently in support of a positive effect of unions on management, it is important to emphasize, for

the purposes of this study, that SHL documented numerous negative effects of unions on

management. For example, they cite a narrowing of the scope of managerial discretion in the post-

unionization phase.

For a better understanding of union impact on management, it is important to frame the shock

effect within the context of a unions monopoly and voice roles (Freeman and Medoff, 1979).


5/48

5

According to SHL, management responds by seeking higher efficiency in response to unionization-

induced shock. One way to obtain better efficiency is to make changes in areas over which

management has exclusive control. We can call this the pure shock effect, i.e., one in which the

union is not directly involved in obtaining new efficiencies. For example, SHL cite the development

of management by policy in contrast to an ad hoc approach and the introduction of changes in

management structure (p. 951-2). These findings point to the standardization and formalization of

managerial workplace practice that takes place as result of unionization but without the direct

involvement of the union. Such changes are management initiated, e.g., a tightening of quality

standards or reduction of waste and may be implemented in many instances without consulting the

union. To summarize it in their own words, The challenge that unions presented to management

has, if viewed broadly, created superior and better balanced management, even though some

exceptions must be recognized (SHL, p. 951).

However, management desire to obtain efficiencies is not limited to policies under

managements exclusive control. There are other areas where changes could be made for higher

efficiency but because those changes would impact on working conditions management would need

to engage the union in this process. Moreover, for unions giving workers a voice in workplace

matters is a primary goal. Therefore, unions are eager to seize every opportunity to question and

modify managerial decisions, when ever possible. Building on Hirschmans concept of voice (as

opposed to exit), F-M suggest that union goals include giving their members a voice in workplace

decisions. At this juncture, a dialogue between union and management over efforts to increase

efficiency can end in a bitter fight settled only by the exercise of their relative power or it could

result in mutual accommodation through constructive give-and-take. Changes brought about by

either of these two processes may be called the voice effect of unions on management. Descriptions

of voice mechanisms that lead to positive union effects can be found in the SHL study. Theoretical


6/48

6

explanations for why such voice mechanisms would contribute to higher efficiency in the workplace

can be in Metcalf (2003) and Hirsch (2004).

The SHL study found that most union-management relationships began with a collective

agreement in the form of a brief, simple document. Over time and with successive bargaining

rounds, these documents became longer and more complex. SHL interpreted this process as a

positive development that yielded fully developed contracts (SHL, 1960, p.186). These contracts

contained a close meeting of the minds in key areas such as discipline for wildcat strikes and a

general decline in conflict over the interpretation of clauses in the collective agreement. This process

of gradual convergence between union and management views of the workplace could have resulted

from mutual and gradual accommodation or it may have involved the use of unions coercive

monopoly power. The SHL study shows that frequent dialogue, both formal and informal, played a

major role in bringing about a convergence in union and management views.

Although F-M make a clear conceptual distinction between the monopoly and voice roles of

the union, it is operationally very difficult to separate the two. Many changes in the SHL study are

reported to come out of dialogue and exchange, but it is not clear that the use of coercive monopoly

power did not contribute to observed union effects. When union-shocked management goes to

workers to obtain additional efficiencies, sometimes management discovers that they can do better

by engaging in dialogue, negotiation and persuasion than by coercion and mistrust. At other times,

either or both parties may use their coercive power as well as dialogue to achieve certain changes in

workplace policies. This suggests that monopoly power and voice, may not always be independent of

each other. For example, a union with greater monopoly power may also enjoy more effective voice

or alternately, a union with low levels of monopoly power may not be so influential in its voice role.

This possible interaction between the two roles makes it hard to empirically separate the pure effect

of monopoly power and voice on management (Hirsch, 2004). Thus, the discussion of empirical


7/48

7

research that follows in the next section can not (and does not attempt to) isolate the pure, or sole,

application of either monopoly power or voice. Rather, it argues that union effects on management

result from the application of both monopoly power and voice.

Theoretical explanations for the existence of a positive union effect. What is less clear from

earlier empirical studies, including SHL, is a theoretical explanation for why unions could have a

positive effect on workplace practices. If markets were perfectly competitive, firms would be able to

arrive at an optimal mix of HRM and workplace policies without any union presence or pressure.

Clearly, this theoretical prediction is not consonant with the evidence from SHL, F-M and some

other studies. Unions do appear to force (or/and persuade) management to adopt efficient practices

that management would have ignored otherwise. An explanation can be offered in terms of

management failure or market imperfections or both. Management failure results, at least partially,

from ambiguities in short- versus long-term goals. Depending on how these goals are defined,

management may focus on some policies and not on others. A good example would be investment in

training which may receive a bigger emphasis in organizations with longer-term objectives than in

organizations with short-term objectives. Similarly, market imperfections such as limited

information, small number of buyers and sellers and coordination failures can lead to the firm

adopting non-optimal use of certain workplace policies. Kaufman (2001) uses a similar approach to

show why firms would adopt less than the optimal amount of worker participation and representation

in the workplace (p. 505-508). Similar arguments can be developed in respect of a range of

workplace and human resource management policies.

One of the much discussed explanations for a positive union effect can be found in

Leibensteins (1966) concept of X-inefficiency which is hypothesized to result from the firm being

prone to producing above the minimum possible costs or above the efficient production frontier. In

this view, X-inefficiency results from firms not employing least-cost combinations of labor and


8/48

8

capital or from not utilizing the factors of production in the most efficient manner possible (Hirsch

and Addison, 1986, p.188). There are many factors that can lead producers to be X-inefficient:

incentives for management, organizational structure, and supervision, among other factors

(Leibenstein 1978; ). Once the union makes its appearance, it shocks management into finding

efficiencies that were hitherto not fully tapped. Of course, this explanation is not without

controversy. Some observers have complained that X-inefficiency is an ambiguous concept (Stigler,

1976). It does not clearly layout a theory of management motivation in which management would

willingly exclude the most efficient methods. However, the X-inefficiency explanation has received

support from both economic arguments and organizational evidence.

Altman (2001) develops a theoretical argument for the existence of a positive union effect

by building on the concept of X-inefficiency. Altmans principal argument rests on the premise that a

range of wages may correspond to the same level of unit cost of production. Although unions push

up wages, it does not mean that higher wages make the producer a higher cost producer (p.104-106).

With union-induced higher wages, the marginal revenue product (i.e., wages X marginal product of

labor) line shifts outward because both workers and employers adopt practices, sometimes by

cooperating together and at other times by forcing the other party to make concessions, to increase

the marginal product of labor. As long as wages rise by an amount equal to the average (not

marginal) product of labor, there should be no negative impact of higher wages on employment. This

explanation for observing a positive union effect is consistent with the concept of X-inefficiency.

Econometric evidence consistent with a positive effect of unions on outcomes such as

productivity has been well documented by F-M. In addition, some organizational studies provide

qualitative evidence to support the explanation offered above. Such evidence suggests that it is very

hard for management, a hierarchical organization, to come up with the most efficient process on its

own because of its in-built inability to question hierarchy or the dominant paradigm. The implication


9/48

9

is that when unions enter the scene they are able to question management. Such questioning sets up a

dialectic, otherwise absent from managerial deliberations, which then leads to better, more creative

and hence, more productive solutions. Rubinstein and Kochan (2001) in their study of the Saturn car

plant cite from the notes of a colleague, Bob McKersie, who sat in on many deliberations of labor-

management interactions:

.. it is clear that the role of the UAW partners is absolutely pivotal for the functioning of Saturn. At the most recent meeting of the SAC [ joint labor-management body ], the only individualswho were willing to take issue and to tell it like it is were the UAW representatives. Other

participants in the meeting did not speak their minds as freely and tended to back off when the CEOexpressed a point of view

(p. 36; italics added )

Although this example by itself is not conclusive, it is consonant with a significant theme in

industrial relations literature that the principal role a union is to question management decisions. A

related dynamic of labor-management interaction may be called the learning effect , i.e., both sides

learn of new arrangements that can be used to govern the workplace and to guide efficient

production. Such learning would be less likely to take place in the absence of unions and the

dialectic they set up in their wake. Clearly, the learning effect, as defined here, would form a part of

the voice effect.

From the foregoing discussion it appears appropriate to conclude that unions have both

positive and negative effects on workplace efficiency. It has been common in the past to label the

positive effect of unions on organizational productivity or efficiency as the shock effect. However,

effects other than management being shocked may contribute to the positive impact of unions, e.g.,

the voice role of unions in the workplace. It is also simplistic to attach positive and negative labels to

monopoly and voice effects respectively. It has been shown above that both monopoly and voice

roles of the union contribute to positive and negative effects on efficiency. Some monopolistic

behavior of the union can have positive outcomes for the organization just as some voice functions

can be inefficient for management. Shock effect can be thought of as the sum total of the positive


10/48


11/48

11

Observing the union effect. Lastly, we need to consider how best to observe the union effect

if indeed there is an effect. The SHL (1960) study was close enough to the first large-scale, enduring

wave of unionization to be able to study management practice before and after unionization using

recall interviews with key informants. This approach may not be possible in many places today.

A direct way to observe the union effect may be to examine workplace policies and practices

before and after the event. Such an event study approach can be useful in many situations but it is

not particularly well suited to unionization where the full impact becomes discernible only after a

few years. Moreover, we have not witnessed any large waves of unionization in the last twenty-five

years in North America; this approach would cover only a small fraction of workplaces.

Another way to measure union impact is to directly compare union and nonunion workplaces

after applying appropriate controls for effects of size, industry, etc. In the U.S., declining

unionization has reduced research interest in such comparisons. Many studies, e.g., periodic surveys

of organizational practices in human resource management by the Centre for Effective Organizations

at the University of Southern California (Lawler, Mohrman and Benson, 2001), do not report their

results by union status. In European studies, this differentiation is frequently omitted because

collective bargaining coverage can be high even when union density is low. Many European workers

and organizations are covered by terms of collective agreements either by law (as in France) or by

industry-level understandings among employers (as in Germany). A direct union-nonunion

comparison in this context would underestimate the true union-nonunion difference.

III. Union Impact on Management

Research evidence on union impact can be divided into three groups. One group of studies points to

the lack of flexibility in unionized systems. For example, unionized workplaces are associated with

wage compression and fixed wages. Nonunion systems on the other hand, are more likely to adopt a


12/48

12

variety of incentive and contingent pay systems. The second group of studies tends to emphasize the

efficiency that comes as a result of union presence, i.e., through formalization and standardization of

management policy. For example, unionized workplaces tend to have more training and other formal

systems such as safety policies. A third group finds no significant difference in HR practices

although many of these studies focus primarily on newer, innovative practices such as employee

involvement and flexible work arrangements, among other such practices. As suggested earlier, a

finding of no differences across the two sectors may be indicative of diffusion of key innovations

from one sector to the other.

The rest of this section reviews empirical evidence from a number of studies that examine the

union-nonunion difference descriptively or estimate the union impact analytically on a range of

human resource management and related workplace practices. This review is organized by various

workplace practices covered by these studies: recruitment and selection, flexible staffing, training,

employee voice, teams and job flexibility, job evaluation, promotion and performance appraisal, pay

systems, overall human resource strategies and quit rates as one example of individual and

organizational outcomes that is closely watched by all parties.

Recruitment & Selection. Koch and Hundley (1997) examined the impact of unions on

recruitment and selection practices using data from a survey originally conducted at Columbia

University of executives from 7,765 business units contained in the 1986 Compustat II Industry

Segment files. The final sample included useable responses from 495 executives. This study

analyzed two sub-samples, one covering all industry groups, and another restricted to manufacturing.

Results from this analysis indicated that in the all-industry sample, unionized firms were

more likely to employ fewer methods of recruitment, such as newspaper, agencies, referrals, and

walk-ins, than nonunion firms. The only exception was in the case of government agencies, where


13/48

13

both union and nonunion firms were equally likely to use this channel. In addition, with partly

unionized firms, the number of recruitment methods significantly declined with the increase in the

degree of unionization of the firm. Partly unionized firms were also found to use fewer recruitment

methods for union jobs than for nonunion jobs. However, as explained in a later section, unionized

firms showed a higher tendency to use formal employment tests. Although these two effects are

seemingly opposite in nature, i.e., one reducing management flexibility while the other encouraging

adoption of more rigorous testing, both effects are consistent with union need for greater

formalization of management decision-making.

In other areas of HRM, Koch and Hundley (1997) found mixed support for the effect of

unions on selection methods , such as skill, aptitude, drug, and physical tests. For example, they

found only mild support for the proposition that unionized firms were more likely to use a larger

number of selection methods. This effect was positive and statistically significant for the all-

industries sample, and positive but not statistically significant for the manufacturing sample.

Further, these researchers found that within a single firm, there appeared to be a similarity of union

and nonunion selection methods indicating the strength of the threat effect with respect to selection

methods. Essentially, the impact of unions appeared to increase the likelihood of the use of only

drug tests and physicals and not skills and aptitude tests, but overall unionized firms have an

increasing tendency to use formal employment tests.

Ng and Maki (1994) report on a comprehensive study of union-nonunion differences in

various human resource management practices using survey data from a sample of 356 organizations

across a variety of industries including Food, Furniture, Fabricated Metals, and Electrical/Electronic

Equipment to examine the impact of unions on. Using multivariate analysis to control for a variety of

factors, they examined the impact of unions on a total of 37 HRM practices from hiring policies to

training to promotion practices. With respect to hiring and recruitment practices, they found that the


14/48

14

presence of a union increased the likelihood that a firm would formally post jobs internally for open

competition among current employees. In addition, unionized firms were more likely to impose

formal probationary periods on new hires. However, regarding external recruitment practices, such

as employee referrals, external job advertisements, and walk-ins, as well as the amount of previous

job experience required of new hires, this study reported no significant difference. In a study of the

auto parts industry, Kaufman and Kaufman (1987) found that unionized plants were nearly twice as

likely to post jobs internally as their nonunion counterparts.

The evidence shows that unions appear to insist on promotion-from-within and the related

use of internal posting-and-bidding. This, in turn, causes management to limit its channels of

external recruiting, and to some extent, use only physical tests in selection. These findings are

consistent with the internal labor market view of organizations (Doeringer and Piore, 1975).


15/48

15

Flexible Staffing. Gramm and Schnell (2001) investigated the use of flexible staffing

arrangements in core jobs and how these arrangements affect job security of regular core employees.

The researchers used their own survey ii to collect data from a random sample of human resource

managers in Alabama establishments that in part employed flexible staff arrangements. Due to cost

saving associated with using flexible staffing arrangements (FSAs), Gramm and Schnell argued that

organizations pursuing low-cost production strategies would be more attracted to this method of

employment. The authors hypothesized that companies that used FSAs to complete core tasks have a

greater ability to adjust to temporary decreases in the demand for labor and therefore are less likely

to layoff regular core employees. The role of unions is important in this analysis. Gramm and

Schnell point out that on the one hand, unions act as a barrier to the dismissal of workers without just

cause which in turn, increases the desire of the employer to use FSAs to gain control of the cost of

labor. Theoretically, the lower costs provide job security for regular core workers by insulating them

from changes in the labor market.

However, unions are often opposed to alternative employment arrangements and through

collective bargaining, i.e., the use of monopoly power, insist on limiting the use of such labor. The

results of their logistic regression indicated that increases in percentage of union representation

decreased the likelihood of the use of FSAs. These findings support the view that unions reduce

management flexibility even though the reduced flexibility may jeopardize job security for their

members in the longer-run.

In addition, Gooderham and Nordhaug (2000) report that firms that perceived their trade

unions as being powerful were less likely to implement HR strategies to increase numerical

flexibility. Various indicators of numerical flexibility include the use of temporary employees, part-

time employees, and subcontracting. In fact, trade union power was found to have a consistently

more powerful effect on the indicators of numerical flexibility than the level of competition. The


16/48

16

researchers intent was to examine factors that affect a firms use of staffing and HR strategies

designed to improve their flexibility in response to the environment in which they operated.

In their study of the auto parts industry, Kaufman and Kaufman (1987) found several

instances of work rules in unionized plants that prohibited some workers from doing the job of other

workers. Between 25-33% of the plants reported prohibiting production workers from doing the job

of other production workers either usually or occasionally. All the nonunion plants said that they

never prohibit workers from doing the job of another. Among the unionized plants, only 43.8% of

the sample could make the same claim. Since this study was carried out many unionized workplaces

have negotiated newer work rules. Thus these estimates of union-nonunion differences may be

somewhat higher than the situation in 2004 at the time of this writing.

Training. Osterman (1995) examined the way American firms train their employees and

whether the type and amount of training can be explained by a firms organizational structure. The

purpose of this study iii was to determine the utilization and distribution of skill levels in a cross-

section of U.S. firms and to quantify the amount of training received by workers at American

Companies.

Osterman estimated the effects of organizational structure on the percentage of core

employees who received formal off the job training. He argued that unions could be a positive

influence on the amount of training received by employees by pressuring establishments to invest in

their workers, or conversely act as an obstacle to increased training with an insistent stance toward

the protection of traditional job rights (p.138). The results of the model showed that the presence of a

union significantly increased the likelihood of a core worker receiving off- the-job training. From

this study, it is hard to infer whether the positive effect on training bis the result of voice or

monopoly power.

Using the British Household Panel Survey (BHPS) covering the period of 1991-1995 and


17/48

17

with an estimated sample of 2982 men and 3117 women, Arulampalam and Booth (1998) examined

the intersection of work-related training and labor market flexibility using type of contract, part-time

employment, and lack of union coverage as a proxy. Using a probit model, the researchers found that

nonunion workers as well as workers covered by short term contracts and part-time workers were

less likely to be involved in work-related skills training. In addition, nonunion males were 7% less

likely to receive training than union-organized males and nonunion females were 10% less likely to

receive training than their unionized counterparts.

Similar inferences on training can be drawn from a study using data from AWIRS 1989-

1990. Kennedy, Drago, Sloan and Wooden (1994) examined the effect of unions on the extent of

formal training provided by employers. The random sample of 2004 workplaces covered firms with

a minimum of 20 employees; interview questionnaires were distributed to management and if

applicable, union delegates. Kennedy et al.s results indicate that the union impact on formally

delivered training programs is positive, but only where unions are active in the workplace. In

addition, active unions also appeared to have a positive net effect on the amount of external training

received. The researchers attributed this finding to the possibility that active unions forced firms

to increase the amount of external general training to improve the overall skill development of the

employees, or that unions encouraged firms to use general training as a fringe benefit. Further

analysis of the results indicated that the connection between tenure and firm-specific training was

strengthened in the presence of an active union. Active unions appeared to reduce the association

between firm size and in-house training. Kennedy et al. attributed this finding to the belief that

unions have a negative impact on efficiency.

In another study of training practices, Heyes and Stuart (1998) gathered data drawn from two

surveys conducted between February and October 1994 and in early 1995 iv. The survey covered the

following topics: 1) attitudes towards training, 2) experiences of vocational and non-vocational


18/48


19/48

19

In a Canadian workplace survey of training practices, Betcherman et al. (1994, 1997), found

that union status of a workplace was positively associated with higher levels of job-related training.

These findings are not unusual and together with other findings about the workings of unionized

workplaces suggest that unions do exert a shock-like effect on management.

Job Evaluation . Ng and Maki (1994) examined the adoption of job evaluation techniques

across union and nonunion firms and found that both types of firms were equally likely to use a

classification or point system method of job evaluation. However, nonunion firms were more likely

to employ more subjective evaluation criteria, while unionized firms employed more objective

ranking evaluation such as the benchmark method. While management may find the use of

subjective criteria suiting its purposes, unions insist on objective criteria in order to reduce

favoritism and to maintain its ability to question management decisions.

Promotion & Performance Appraisals. In respect of promotion procedures, the Ng and Maki

(1994) study also found that unionized workplaces were more likely to formalize promotion

procedures than nonunion firms. This finding fits the pattern of union preferences in which they want

management to formalize all procedures by writing them down in some form. It is in the

formalization of procedures that the union ability to question management lies. The seniority rule is a

highly formalized rule that may be inefficient for management but it gives the union considerable

power in challenging managerial decisions.

In their study of the auto parts industry, Kaufman and Kaufman (1987) found that seniority as

the sole basis for promotion was not common among either union or nonunion plants. However, if a

worker could do the job then seniority became the decisive factor in many more unionized plants

than in nonunion plants. Ability as the deciding factor in promotions was much more prevalent in

nonunion plants than in unionized plants. The same pattern of differences were reported in the case

of layoff decisions but the union-nonunion difference was statistically not significant.


20/48

20

In respect of performance appraisals, Ng and Maki (1994) found unionized firms to be less

likely to employ a formal appraisal system. This finding can be related to the way performance

appraisal data are used in the organization. Regarding specific use of performance appraisals data,

the regression results indicated that while both union and nonunion firms were equally likely to

employ the results of an appraisal in disciplinary and training decisions, unionized firms were again

less likely to use appraisal results in salary, promotion, and layoff decisions. For unions, the

efficiency effects of performance appraisal systems are not large enough to overcome the costs of

loss of solidarity among its ranks. Since unions generally view management-run performance

appraisals as subjective processes that serve management purposes at the expense of union goals, it

is only logical that they oppose such systems or at the margin, limit their applicability to decision-

making.

Pay, Variable Pay, and Incentives Systems. There is fairly widespread and robust evidence

that the presence of a union in the workplace greatly reduces the likelihood of a variable pay plan

(VPP) being used. Betcherman et al. (1994) found that VPPs such as profits sharing, ESOPs,

knowledge pay, merit pay, etc., were more often employed in non-union firms. The overall incidence

of VPPs is close to 50% higher in workplaces without a union. They found that the only VPP that

more unionized firms than non-unionized firms implemented was productivity gainsharing, and that

figure was only marginally higher than in non-union plants. In the case of other benefits, the study

did not find any significant difference in family-care benefits between unionized and non-union

firms, except for the incidence of employee assistance programs (EAPs) which were higher in

unionized firms. Nearly 45.7% of the unionized firms offered EAPs, while only 28.8% of non-union

firms provided them.

Further support for this finding comes from the recent Canadian Workplace and Employee

Survey (WES) which since 1999 has surveyed over 6500 workplaces and 25,000 employees each


21/48

21

year. Unionized workplaces are less likely to use to individual incentive plans compared to nonunion

workplaces (Verma and Fang, 2002). This is not true of group incentive plans, which were found to

be occurring with very similar frequency in both union and nonunion workplaces.

This survey showed that unionized firms have a much lower incidence of individual incentive

plans than nonunion firms. Only 13.6 % of unionized firms used incentive compensation systems

such as merit pay, profit sharing, and straight piece rate plans, while 28% of nonunion firms had

implemented a profit sharing plan. Kaufman and Kaufman (1987) also report that nonunion plants

were nearly nine times as likely to adopt a profit sharing plan as unionized plants. When we put this

evidence together with stated union policy against pay systems that compensate workers individual

workers differentially on the basis of management-assessed performance criteria, it is logical to infer

that the difference in pay policies of unionized firms reflects union desire to adopt uniform pay

rather than pay-based variance in individual performance. Interestingly, as pointed out in a following

section, there is no significant difference between unionized and nonunion firms in their adoption of

pay based on group performance that can be measured and verified directly. This criterion

distinguishes profit sharing plans from gainsharing plans. Profit sharing is much more likely to be

found in nonunion workplaces because unions are generally opposed to profit sharing. Their main

objection appears to be rooted in the fact that accounting profits can be volatile over time and only

loosely connected to workplace effort and productivity. Gainsharing, on the other hand, is more

likely to be found in unionized workplaces, because it is generally based on verifiable group

performance at the workplace level.

In another study, using interview and survey data collected over a five-year period from 53

large (greater than 100 employees) greenfield sites in Ireland, Gunnigle et al. (1998) examined the

implications of performance-related pay (PRP) systems on the collective aspect of the labor

relationship. The researchers found that the use of performance pay based on individual performance


22/48

22

appraisals was highly negatively related to the presence of a union. Specifically, nonunion

companies were significantly more likely to use PRP systems. In fact, 75% of the nonunion firms

used performance pay systems based on individual performance appraisals. Gunnigle et al. suggest

that by allowing managerial discretion over the individual performance appraisal, nonunion firms

could employ individualist as opposed to collectivist incremental pay decisions, thus challenging the

collectivism of a unionized environment and collective bargaining.

Some studies suggest union effects even when the union status variable was not directly

examined. If unions push for policies such as time-based pay systems and if it is found that such pay

systems reduce organizational flexibility, then we may infer that unions have some impact on

organizational flexibility. In one such study, Long (2001) examined the relationship between a firms

pay system and its ability to achieve structural flexibility in a study of 44 large Canadian

manufacturers including forest products, hard goods, high technology, petroleum, pharmaceuticals

and mining. The study examined structural flexibility that includes having a) informed employees,

that have b) training and knowledge, who are c) empowered to make decisions, operating under a d)

reward system that creates common goals. The types of pay systems examined included: group pay,

such as profit and gain sharing, time-based pay, seniority pay, and individual performance pay.

Results showed a significant and positive relationship between group pay and flexibility and

a significant negative relationship between time-based pay and flexibility. In addition, multiple

regression results indicated that pay system-related variables explained a substantial amount

(approximately 20%) of the variance in structural flexibility. Long concluded that the results of the

study to appear to provide some support that the type of pay system used by a firm may affect that

firms structural flexibility. We know that unionized firms are more likely to adopt time-based pay.

The findings of this study suggest that unionized firms can be expected to experience lower levels of

organizational flexibility.


23/48

23

Although unions generally discourage firms from adopting individual pay incentives, they

appear to be less averse to group incentives based on directly measurable and verifiable criteria. In

the Ng and Maki (1994) study, both union and nonunion firms were found to be equally likely to

adopt group level incentive pay systems and gain sharing plans. This suggests that even though

unions have a different preference pattern in many areas of human resource policies, there are some

areas where their preferences are not that different from managements. This type of convergence

between HR policies of union and nonunion firms is in conformance with the notion of innovations

diffusing from one sector to another.

Although gainsharing is relatively more popular with unions, some studies have found a

mixed picture in terms of the impact of unionization (Kim 1996). Kim explored the relationship

between gainsharing programs and employee involvement to find that unions seem to reduce the

effectiveness of gainsharing plans even as they marginally improve their survival. The study v

focused on the overall effectiveness of compensation schemes that included gainsharing and the

factors that increased the likelihood of a successful gainsharing program. Included in the analysis

were: a measure of unionization, i.e. union presence in the organization, and union support, i.e.,

whether the union supported or opposed gainsharing.

In regard to the union variables, results of the ordered probit analysis indicated that

gainsharing programs in unionized settings were significantly less successful in influencing

organization performance in terms of improved quality, improved labor productivity, cost reduction

and improved production process relative to establishment without unions. There was no significant

difference in bonus payouts. However, results from a reduced sample of only unionized companies

indicated that when there was support for gainsharing the program was significantly more likely to

yield positive results of quality, cost reduction, improved production process and bonus payments.

In another study, Kim (1999) examined the survival of gainsharing programs at 211 different


24/48

24

organizations in the United States and Canada. The data used in the analysis was taken from a survey

collected in 1992 (see Kim 1996). As mentioned earlier, unionization increased the likelihood of a

gainsharing programs survival, all else being equal, to a small extent. Organizations that utilized

outside consulting in the development and maintenance of their gainsharing programs were more

likely to see them fail and disappear. Employee approval, new employee training, customization,

financial performance and capital investments signified successful gainsharing programs. While

organizations in the manufacturing industry had significantly less success with their gainsharing

programs, unionization did not significantly affect the success of these programs.

Organizational Climate and Workplace Culture. Unions impact organizational climate and

workplace culture in a variety of ways. Since the full scope of this topic is vast, only selected aspects

are presented here. One set of studies point to a more formalized and structured communication

between management and employees and a related loss of informality in unionized workplaces. On

the other hand, managements in nonunion environments are able to build a climate of effective

communication and strong identification with organizational goals. In one qualitative study, Cohen-

Rosenthal and Burton (1993) found that unions were a barrier to effective communications with

management. By opposing the HR manager frequently, the union can often cloud the intended

message from management. By acting as an intervenor, the presence of a union can impede and

increase the costs of direct communications between management and the employees. Many

nonunion employers can achieve greater clarity in communicating with their employees because they

are able to establish direct links with the employees.

Citing an earlier study which employed both surveys and interviews to a sample of 248

workers from three large nonunion and four large unionized U.S.-owned plants in Ireland, Flood and

Toner (1997) examined whether any advantage can be derived from nonunion firms following best

practices human resources strategies or otherwise, nonunion firms that practice union avoidance


25/48

25

strategies. Using multivariate analysis, Flood and Toner concluded that the presence of a union may

decrease a firms ability to design and adopt human resource policies that can assist in increasing

employee motivation and cooperation, and building a strong corporate culture. In addition, the

researchers found no evidence that nonunion status increases the firms overall flexibility; however,

the nature of the industry needs to be considered as more traditional industries may have a history of

inflexibility.

Taras (2000) examined the level of employee representation in the joint industrial council

(JIC) that operates as an enterprise level employee organization at Imperial Oils upstream division.

Through observing meetings, reviewing corporate minutes and in-depth interviews from 1995 to

1999 vi, Taras found that Imperial Oil and their JIC employed numerous collaborative and

competitive HR techniques that are demonstrative of the union threat effect. For example, the

primary responsibilities and scope of JICs at various levels cover three areas of workplace policies:

competitive wages, benefits and working conditions, strengthening corporate culture through

employee affiliation, and providing employee voice through committees and sub-committees. It may

be inferred that one of the reasons that management continued to work with the JIC, even though

they did not have to do so by law, is for the positive contribution of the JIC to the culture and climate

of the workplace.

Other ways in which Imperial Oil contributes to the relationship is by allowing hourly

employees to have access to typically confidential information such as competitors compensation

structures and other HR-related information. Although the JIC may have little power in influencing

actual wages and benefits structures, the various JIC representatives meet collectively and share

information, increasing employees sense of corporate affiliation. This sets the tone for a positive

climate.

Regarding the creation of an integrative corporate culture, the JIC enables a work site to vote


26/48

26

on any workplace issues, except those involving corporate policy. Other recent union-competitive

issues that the JIC has been involved in have been downsizing and job elimination. In fact, Imperial

Oil states that the JIC operates similarly to a union in providing employee voice as it also oversees

discipline and grievance procedures; JIC representatives get involved at the initial stages and remain

involved until the final grievance steps.

The Australian Workplace and Industrial Relations Survey (AWIRS) conducted by the

Department of Industrial Relations in Australia in 1990 and 1995 is also helpful in understanding the

union-nonunion difference in organizational climate (Callus, 1991; Morehead, 1997). Among other

findings, the AWIRS indicated that more workplaces had adopted a modern workplace relations

style, implying a decrease in the active role of unions coupled with an increase in a more structured

approach by management. This suggests that the presence of unions tends to decrease

managements ability to structure workplace practices and thus, as the role of unions has been

diminishing over time, managements are able to take a more active role in structuring workplace

climate and culture to their perceived operating needs.

Of the managers surveyed in AWIRS, who indicated they wanted to make changes in pay

systems (51%), 17% stated that union resistance acted as a barrier preventing them from

implementing those changes. Although the degree of unionization declined over the 5-year time

frame, AWIRS results indicated that union delegates became more actively involved negotiating key

employment issues such as wage increases, perhaps indicating a more communicative relationship

between union delegates and management. A rise in the incidence of grievance procedures and joint

consultative committees over the same timeframe appears to support this statement. The AWIRS

study also found that higher than average numbers of unionized employees felt that the changes that

had occurred in their workplace over the last 12 months had negatively affected them and thus, could

be considered worse off. In examining employees attitudes towards trust and satisfaction,


27/48

27

unionized employees were also less likely than non-union employees to indicate satisfaction and

trust with management. Therefore, although different factors may affect both managers and

employees level of satisfaction and trust, the presence of unions appears to negatively affect the

type of management relationship and attitudes in the workplacevii

.

In the area of workplace safety and health, another important contributor to organizational

climate, Weil (1991, 1992) found that unionized workplaces were much more likely to have labor

inspections than nonunion workplaces even when the safety and health legislation applied equally to

both union and nonunion workplaces. The differential inspection rates are rooted in a climate of

safety created by union presence. Clearly, unions play a role in the labor inspection process, which in

turn impacts management. In a similar vein, Beaumont and Harris (1996) study using UK 1990

WIRS data suggests that management motivation for the introduction of HRM practices often arises

from dissatisfaction with and a desire to improve the existing labor-management relationship.

Taken together, these findings suggest that unions generally reduce the scope of management

prerogatives and discretion in managing the workplace. It is harder for management in the presence

of a union, to foster among workers a strong identification with corporate goals and vision. The

confirmatory evidence for this inference lies in the evidence from many nonunion workplaces where

managers are only too willing to offer workers greater voice, better wages and benefits. In the

absence of a union, many managers also appear to develop policies to foster a strong corporate

culture that would persuade workers that their interests are the same as the organizations.

Employee Voice and Communications. There is a considerable body of indirect evidence that

suggests that unions provide a significant amount of voice. As argued earlier, some of the best

though indirect, evidence of union impact on management comes from an examination of

management policies in the nonunion sector. In this section, a number of studies are reviewed that

report on voice mechanisms in the absence of a union. While union threat can be a major factor in


28/48

28

adoption of employee voice and communications by nonunion firms, it is important to note that this

is not the only reason for which firms would adopt these practices. Other reasons would include a

desire to improve efficiency, improve employee satisfaction and a desire to avoid costly law suits

that may result from inappropriate employee discipline and discharge in the absence of employee

voice.

In a sample of 18 cases from the UK, Gollan (2000) found only 11% of the workplaces to

have a representative committee. Of the committees that do exist, these nonunion employee

representation plans (NERs) discuss potentially threat effect-induced issues, such as pay, basic work

conditions, hours, staffing levels, new technology organization, and the development of new

products and services. However, Gollans evidence indicates that few committees have negotiation

and bargaining rights; thus they are not as effective at affecting policy change. In addition,

managers, not NERs, are primarily responsible for resolving grievances and conflict resolution.

Thus, NERs appear to have limited ability to influence wages, policy and strategic issues, and how

and when workplace changes are introduced. Aside from these limitations, NERs in the UK and

Australia appear to represent good communication mechanisms but are less effective at influencing

other substantive HR policies.

The evidence from a Japan-based survey viii also provides similar evidence on possible

indirect effect of unions on HR practices in the nonunion sector (Morishima and Tsuru, 2000).

There is a wide range of non-union representation plans in Japan, e.g., some non-union firms work

with employee representatives, while other nonunion organizations are more similar to trade unions.

In addition, employee associations are also called friendship societies or staff councils, perhaps an

attempt to indicate a type of collaboration or affiliation between the employer and employees. These

bodies regularly discuss issues such as wages, benefits, health and safety and other issues over which

unions bargain with the employer. Thus, effectively, many rights of representation are extended to


29/48

29

these nonunion workers in response to the threat of unionization.

Morishima and Tsuru (2000) compared the differences between individual and collective

voice mechanisms between union and nonunion firms. In this Japanese sample of firms, significant

differences exist between the provision of collective and individual voice between union and non-

union workplaces. Not only have nonunion Japanese firms attempted to match direct union

workplace benefits such as compensation structures, etc., but have also succeeded in providing

indirect benefits such as a wide range of voice mechanisms beyond that of union firms. On several

indicators of collective voice such as communicating the companys strategy, discussions with

management at both senior and middle levels, employee associations and conducting opinion

surveys, nonunion firms lead unionized firms. Of course when it comes to formal voice mechanisms

such as joint consultation systems and grievance procedures, many more such programs are found in

unionized workplaces. Unionized workplaces were also marginally more likely to include (53.8% vs.

47 %) forms of written reports at the individual level. This difference is not that great and is likely a

product of greater formalization of practices in the unionized sector.

Several studies have found that unions have a positive impact on employee voice

mechanisms. In an Australian study, Benson (2000) found that unionized workplaces were more

likely than nonunion firms to implement employee voice mechanisms. Specifically, unionized firms

were more likely to use collective negotiation, employee surveys and meetings, grievance and equal

employment procedures, consultative and safety committees, task forces, and health and safety

representatives. Other voice mechanisms typically associated with HRM strategies, such as

individual employee negotiations, involvement and consultation) were evenly distributed across the

nonunion, inactive and active union workplaces. Yet, nonunion workplaces were still less likely than

union workplaces to use many of the HRM-associated voice mechanisms, specifically, quality

circles, employee surveys, supervisor-employee meetings, senior management-employee meetings,


30/48

30

and semi/fully autonomous work groups. In addition, the use of quality circles, employee surveys

and supervisor-employee meetings was significantly higher for active union workplaces compared to

nonunion firms.

Benson concluded that the presence of unions appeared to significantly increase the number

of voice mechanisms available to unionized employees compared to that of nonunion firms. In

addition, the more active the union, the greater the number of alternative voice mechanisms used.

More such evidence can be gleaned from a study by Goll (1991) who found that in unionized

settings progressive decision-making positively influenced the number of participative programs and

the number of employees in these programs. These relationships did not exist in the non-unionized

settings.

Using the 1990 Workplace Industrial Relations Survey (WIRS), Fernie and Metcalf (1995)

examined the effects of three forms of employee participation: 1) employee involvement, 2)

contingent pay, and 3) forms of representation on workplace outcomes such as productivity levels

and employment changes, specifically the labor relations climate, quit rates, and absenteeism rates.

Their analysis indicated that compared to a nonunion firm, the unionized firm experienced lower

employment growth and worse labor relations climate. However, the presence of a union was also

associated with lower quit rates. Fernie and Metcalf attributed this occurrence to the unions

provision of a collective voice. Their study also found that the presence of a union had no

significant effect on the rate of absenteeism. Further, joint consultative committees were found to

have only weak or no relationship on these workplace outcomes.

There is a wealth of case-studies of individual firms that document voice systems being

adopted partially in response to the union threat. Delta Airlines, a major nonunion carrier in the U.S.,

has provided employee voice mechanisms since the companys establishment in 1924 (Kaufman,

2003). These actions include providing a competitive work environment, employee stock options, a


31/48

31

profit sharing plan, building strong corporate culture, competitive wages and benefits and the Delta

Personnel Board Council, a management-employee forum used as a communications channel

between the board of directors and employees (Cone, 2000; Kaufman, 2003). Some of Delta Air

Lines leading competitors such as United Air Lines are highly unionized and it is possible that the

union threat keeps Delta from discontinuing employee voice forums. But given the long-standing

policy of providing employee voice it is likely that Delta is motivated by other factors as well.

At the Imperial Oil Cold Lake facility, the company provides employee voice through a

formal employee forum (Boone, 2000). However, the company maintains direct control over

employee-related financial factors, such as pension plan design and compensation philosophy. This

Imperial Oil facility combats the threat effect by giving employees the opportunity to participate in

decision-making, but does not allow for employee input in more financially strategic aspects of the

business operations.

Dofasco, an integrated Canadian steel producer, maintains several voice programs for its

employees in an industry which remains highly unionized. Harshaw (2000) reports that Dofasco has

adopted employee involvement policies that match or exceed workplace practices within the union

sector. For example, profit sharing and employee savings plans were established early on in

Dofascos history. Voice is provided on these topics through a formal representation forum where

employees can provide feedback and input on activities that affect them. Dofasco employs numerous

formal and informal employee representation activities, such as the use of employee focus groups to

solicit feedback on employment policies and practices, and the availability of their company policy

manual regarding the terms and conditions of employment.

Other forms of employee representation include: the companys open door policy with access

to the company president, if necessary, and an emphasis on teamwork as the company has adopted

multiskilled teams that are compensated based on competency pay. The company also attempts to


32/48

32

incorporate employee involvement in most corporate activities, such as recreation, training,

equipment selection, job design, customer interaction, and quality improvements.

IV. A Scorecard for Unions

The empirical findings discussed above are mixed in two ways. First, the evidence suggests that

union effects on management objectives can be positive, negative or neutral, depending on the type

and scope of management policy (for a summary of findings see Table 1). Second, the dominance of

voice or monopoly power mechanisms may be apparent in some instances but in many other

situations it is hard to disentangle the separate contributions of voice and monopoly power because

the two often act together to reinforce each other. The overview of empirical research presented

above can be used to develop a scorecard for unions. In the following section, I discuss areas in

which unions reduce flexibility for management, areas where unions help improve management

practices, areas where unions make little or no significant difference, and lastly, areas where unions

have had an indirect impact on managerial practice.

Unions reduce management flexibility. There are several ways in which unions appear to

reduce management flexibility. For example, union representation has been found to decrease the

likelihood of management use of flexible staffing arrangements (Gramm and Schnell, 2001). Unions

may also contribute to reduced organizational flexibility by opposing policies that could create

flexibility for management (Long, 2001).

More specifically, one may cite union effect on personnel practices such as recruitment and

promotions as examples of limiting the full range of management prerogatives. Unionized firms are

more likely to employ fewer methods of recruitment , such as newspaper, agencies, referrals, and

walk-ins, than nonunion firms (Koch and Hundley, 1997) and are more likely to formalize promotion

procedures than nonunion firms (Ng and Maki, 1994). In the compensation area, there is fairly robust


33/48

33

evidence that the presence of a union in the workplace greatly reduces the likelihood of a variable

pay plan (VPP) being used. Studies by Betcherman et al. (1994), Gunnigle et al. (1998) provide

support for this finding. Since use of variable pay is important to many managers, it would appear

that unions clearly reduce management prerogatives in the area of compensation and rewards. There

are other ways that unions can limit management prerogatives: unions prevent management from

implementing HR policies that could foster a strong corporate culture to facilitate employee

motivation and commitment (Flood and Toner, 1997); and, in many situations, unions may also act

as a barrier to communication between management and workers (Cohen-Rosenthal and Burton,

1993). These empirical findings support the view that unions reduce management flexibility even

though the reduced flexibility may jeopardize job security for their members in the longer-run.

Unions help improve management practices. Unionized workplaces appear to enjoy an

advantage in a number of areas of workplace policy. For example, there is fairly robust evidence

from several countries and over many years that a workplace is likely to offer more training to

workers if it is unionized than if it nonunion (Arulampalam and Booth, 1998; Heyes and Stuart,

1998; Betcherman et al., 1997; Osterman, 1995; Kennedy et al., 1994; Hundley, 1989).

Some recent evidence also corroborates the finding of greater formalization in recruitment

and selection. Unionized workplaces are more likely to use a formal procedure to post jobs (Ng and

Maki, 1994) and to use more objective rather than subjective tests in selection (Koch and Hundley,

1997). Presumably, these practices obtain efficiencies that nonunion organizations forego given their

pursuit of other objectives including their desire to remain nonunion. Lastly, unions appear to

increase the quality and quantity of voice forums in the organization. Many studies document the

positive association between union status and worker voice (Benson, 2000; Goll, 1991) while others

go on to show that as voice increases it reduces the quit rate (Rees, 1991; Fernie and Metcalf, 1995).

The overall picture that emerges from these studies is one of unionized workplaces being


34/48

34

more formal in their adoption of certain HR practices that are known to create efficiencies for

management. These practices include, but are not limited to, training, selection and recruitment,

employee voice and group incentive plans such as gainsharing. Thus, empirical evidence is

supportive of a positive score for unions in these areas of managerial practice.

No significant difference. In an earlier section, it has been argued that union impact on

management policy may not endure over time. If an innovation in the union sector begins to

diffuse through the industry either because it is an efficient practice or due to union threat, empirical

observation may show no differences across the union and nonunion sectors and we may infer that

union effects have dissipated over time.

Several studies find little or no impact of unions on some management policies. For example,

Osterman (1994) found that the presence of unions does not affect the likelihood of a firm adopting

innovative work practices; therefore, neither union nor non-union firms appear more likely to adopt

practices such as teams, job rotation, and TQM, etc. Ichniowski et al. (1989) found no significant

difference between the progressiveness of HR policies across unionize and non-unionized firms.

However, unionized firms were less likely to adopt formal performance appraisals and flexible job

design programs. Finally, Cohen-Rosenthal and Burton (1993) examine the diffusion of advanced

work practices (teams, rotation, TQM, and QC) to find no significant union effect. In all these

cases, union effects that may have existed at one time, had dissipated at the time of these studies.

These developments are consistent with a finding of decreasing union wage premium over time in

the U.S. (Bratsberg and Ragan, 2002) and in Canada (Fang and Verma, 2002).

When a previously known union-nonunion difference disappears over time it could be

attributed to the diffusion of workplace practices over time either due to the innovative nature of the

practice or due to union threat. However, in the absence of historical information about a union-

nonunion difference, it is also possible that a finding of no difference is simply reflecting the fact


35/48

35

there are no differences between the two sectors.

Indirect impact of unions on managerial practice. A significant impact of unions on

managerial practice is observed in managerial choice of HRM policies in nonunion workplaces that

provide employees with union-like benefits, due process and voice. This indirect effect of unions on

managerial behavior can be observed only in the long-run. Cross-sectional snapshots of union-

nonunion differences do not adequately capture managements strategic choices (Kochan, Katz and

McKersie, 1986). If there are union-nonunion differences in managerial practice in the near term, it

is likely that management will incorporate such impacts into their long-term decision-making about

where to invest in the future. In the 1980s and the 1990s a number of studies have documented the

tendency in management to invest away from unionized operations and into greenfield sites that can

be run as nonunion shops (Verma, 1985). HRM practices in such firms are designed in direct

response to the unionization threat. Thus, it is logical to view HRM practices of nonunion firms as a

managerial response to unionism.

There is research evidence that some employers respond to the union threat by offering

employees many services provided by unionized workplaces. The existence of extensive voice

forums in nonunion firms such as Imperial Oil (Taras, 2000; Boone, 2000; Chiesa and Rhyason,

2000), Dofasco (Harshaw, 2000) and Delta Airlines (Cone, 2000; Kaufman, 2003), to name just a

few such firms, point to this indirect effect of unions. Thus, these indirect union effects are not as

strong as the direct effects discussed earlier but our observation of the union impact on management

would be incomplete if we were to ignore these milder, yet significant, effects.

V. Conclusions: Unions and Managerial Practice

A score card for union impact on management policy can be drawn up in a variety of ways.

This paper has first outlined an examination by substantive areas of workplace and HRM policy.


36/48

36

Next, the same results have been summarized according to their impact on management: positive,

negative, neutral or indirect. Almost forty-five years after Slichter, Healy and Livernash (1960)

published their findings of a positive union effect (including the shock effect) on management, many

recent studies continue to find empirical support for contemporary versions of a positive union

effect. Faced by unions independent voice, management tends to develop formal systems and

procedures, many of which contribute to efficiency and organizational effectiveness. To single out

one area where this effect can be seen most clearly and reliably, one would have to point out to the

higher incidence of formal training. This is possibly a positive score for unions. However, a full

assessment needs to consider that some union impacts although seemingly positive (e.g., training

enhances skills and earnings) can hide an inefficient allocation of resources that results from the

price effect of unions.

There is evidence that some aspects of union effects erode over time. The principal evidence

here comes from studies that find no difference between management in unionized and nonunion

operations. The erosion is possibly from both ends: unionized firms find it increasingly difficult to

provide a union premium in a more competitive market while nonunion firms continue to catch up to

(and in many cases lead) practices in the unionized firm. On the whole, this is a negative for unions

as it reduces the union appeal to workers. Unions find it increasingly more difficult to recruit

members based on the instrumental belief that joining a union will lead to gains in terms and

conditions of employment. Some writers have argued that such instrumental beliefs are no longer

sufficient to attract new members and hence unions need to turn to other types of appeals (Kochan et

al., 2004; Verma, Kochan and Wood, 2002).

The evidence also suggests that unions do limit flexibility, at least from the management

perspective. Part of this effect can be attributed to unions need to curb management prerogatives.

But, in many other instances the inflexibility originates in union unwillingness (or footdragging) to


37/48

37

accommodate changes to established practice. Rapidly changing markets or technologies require

faster adaptation. While ideology frequently prevents unions from agreeing to introduce flexible

work practices, it is the difficult task of selling such concessions politically within the union that

accounts for many instances of observed workplace inflexibility. This is a negative score for unions

independent of the underlying reasons. When firms are in financial difficulty, this inability to adapt

to the external environment hurts the image of unions in the eyes of not only employers but also

employees, the government and the public opinion.

To the credit of unions, it should be pointed out that many unions have agreed to introduce

changes. Several studies of change and adaptation in the unionized sector document mutual gains

and adoption (Kochan and Osterman, 1994; Applebaum and Batt, 1994; Cohen-Rosenthal and

Burton, 1993). This would give the unions a positive score except that the small victories are

frequently considered by many to be swamped by the large negative score on the pace and

magnitude of adaptation to the market and new technologies.

Unions are strongly associated with giving workers a voice in the workplace. This is true in

the unionized sector where practices such as formal grievance procedures are almost ubiquitous.

Unions are associated with a variety of voice forms that can include joint committees and

representation at the Board of Directors or other senior-level forums for policy consideration and

deliberation. This counts as a positive score for unions.

Unions can be associated with improving voice in the nonunion sector as well. Over the

years, nonunion workplaces have increasingly adopted joint committees, grievance procedures, focus

groups, employee surveys and other forms of employee voice. The adoption of voice systems could

be driven by factors such as the threat of unionization or the diffusion of best practice. In such

cases, the credit can be given largely to unions. Of course, there are other factors such as the rising

cost of litigation (e.g., in the case of unjust dismissals) that may be forcing some employers to adopt


38/48

38

voice systems such as grievance procedures. In that case, the credit for the diffusion of voice systems

can not be given to unions alone. But, high litigation costs are a bigger factor in the U.S. than in

many other parts of the world. Moreover, grievance procedures constitute only a small fraction of

voice systems, a large share of the credit for diffusion of voice systems in nonunion workplaces can

be given directly or indirectly to the union.

Unions appear to have made a significant impact on management and on workplace HR

practices. However, as shown in this paper, this effect is not always easy to separate from other

effects. The union effect is also not static over time. It varies in response to competitive conditions in

product and factor markets. To answer the question of union effects on management more precisely,

more carefully designed studies are needed that control for the effect of other factors. Until better

results are available, static differences between union and nonunion sectors will remain indicative

but not definitive of the impact of unions.


39/48

39

REFERENCES

Altman, Morris. Worker Satisfaction and Economic Performance: Microfoundations of Success and Failure.Armonk, N.Y.: M.E. Sharpe, 2001.

Appelbaum, Eileen and Rosemary Batt. The New American Workplace: Transforming Work Systems in the

United States . Ithaca, N.Y.: I.L.R. Press, 1994.Arulampalam, Wiji and Alison L. Booth. Training and Labor Market Flexibility: Is there a tradeoff? British

Journal of Industrial Relations 36 (December 1998): 521-36.

Batt, Rosemary, Alexander J.S. Colvin and Jeffrey Keefe. Employee Voice, Human Resource Practices, andQuit Rates: Evidence from the Telecommunications Industry. Industrial & Labor Relations Review 55 (July 2002): 573-94.

Beaumont, Phillip B. and Richard I D Harris. "Good Industrial Relations, Joint Problem Solving, and HRM." Relations Industrielles/Industrial Relations 51 (Spring 1996): 391-402.

Benson, John. Employee Voice in Union and Non-union Australian Workplaces. British Journal of Industrial Relations 38 (September 2000): 453-59.

Betcherman, Gordon, Kathryn McMullen, Norm Leckie and Christina Caron. The canadian workplace intransition. Kingston, Ontario: I.R.C. Press, 1994.

___________, Norm Leckie and Kathryn McMullen. Developing skills in the Canadian Workplace: the Results of the Ekos Workplace Training Survey. Ottawa, ON.: Canadian Policy Research Networks,1997.

Boone, David J. "Operation of the Production District Joint Industrial Council, Imperial Oil". In BruceKaufman and Daphne Taras, eds. Nonunion employee representation: history, contemporary practice,and policy. Armonk, N.Y.: M.E. Sharp, 2000, pp. 457-62.

Bratsberg, Bernt and Jr. James F. Ragan. "Changes in the Union Wage Premium by Industry." Industrial & Labor Relations Review 56 (October 2002): 65-83.

Callus, Ron. Industrial relations at work: the Australian workplace industrial relations survey. Canberra,Australia: Australian Government Publishing Service, 1991.

Chiesa, Rod and Ken Rhyason. "Production District Joint Industrial Council at Imperial Oil Ltd.: ThePerspective from the Employee's Side". In Bruce Kaufman and Daphne Taras, eds. Nonunionemployee representation: history, contemporary practice, and policy. Armonk, N.Y.: M.E. Sharp,2000, pp. 474-76.

Cohen-Rosenthal, Edward and Cynthia Burton. " Mutual Gains: A Guide to Union-Management Cooperation ". 2 nd ed., rev. Ithaca, N.Y.: I.L.R. Press, 1993.

Cone, Cathy. "Delta Personnel Board Council". In Bruce Kaufman and Daphne Taras, eds. Nonunionemployee representation: history, contemporary practice, and policy. Armonk, N.Y.: M.E. Sharp,2000, pp. 469-73.

Doeringer, Peter B. and Michael J. Piore. Internal Labour Markets and Manpower Analysis, With a New Introduction. Armonk, N.Y.: Sharp, 1985.


40/48

40

Fang, Tony and Anil Verma. "Union Wage Premium." Perspectives on Labor and Income 14 (Winter 2002):

17-23.

Fernie, Sue and David Metcalf. Participation, Contingent Pay, Representation and Workplace Performance:Evidence from Great Britain. British Journal of Industrial Relations 33 (September 1995): 379-415

Flood, Patrick C. and Bill Toner. Large Non-Union Companies: How Do They Avoid a Catch 22? British Journal of Industrial Relations 33 (January 1997): 257-77.

Freeman, Richard B. and James L. Medoff. The Two Faces of Unionism. Public Interest 57 (Fall 1979): 69-93.

___________ and James L. Medoff. What do unions do? New York: Basic Books, 1984.

Goll, Irene. Environment, Corporate Ideology, and Employee Involvement Programs, Industrial Relations 30 (Winter 1991): 138-149.

Gollan, Paul J. "Nonunion Forms of Employee Representation in the United Kingdom and Australia". InBruce Kaufman and Daphne Taras, eds. Nonunion employee representation: history, contemporary

practice, and policy. Armonk, N.Y.: M.E. Sharp, 2000, pp. 410-52.

Gooderham, Paul N. and Odd Nordhaug. Flexibility in Norwegian and British Firms: New challenges for European Human Resource Management. St. Martin's Press, 2000.

Gramm, Cynthia L. and John F. Schnell. The Use of Flexible Staffing Arrangements in Core ProductionJobs. Industrial & Labor Relations 54 (January 2001): 245-59.

Gunnigle, Patrick, Thomas Turner, and Daryl DArt. Counterpoising Collectivism: Performance-related Payand Industrial Relations in Greenfield Sites. British Journal of Industrial Relations 36 (December 1998): 565-79.

Harshaw, Mark. "Nonunion employee representation at Dofasco". In Bruce Kaufman and Daphne Taras, eds. Nonunion employee representation: history, contemporary practice, and policy. Armonk, N.Y.: M.E.Sharp, 2000, pp. 463-68.

Heyes, Jason and Mark Stuart. Bargaining for Skills: Trade Unions and Training at the Workplace. British Journal of Industrial Relations 36 (September 1998): 459-67.

Hirsch, Barry T. What Do Unions Do for Economic Performance . Journal of Labor Research 25 (Summer 2004): 415-455.

Hirsch, Barry T. and John T. Addison. The Ecnomic Analysis of Unions: New Approaches and Evidence.Boston: Allen and Unwin, 1986.

Hundley, Greg. Things Unions Do, Job Attributes, and Union Membership. Industrial Relations 28 (Fall1989): 335-355

Ichniowski, Casey, John T. Delaney and David Lewin. "The New Resource Management in US Workplaces:Is It Really New and Is It Only Nonunion?" Relations Industrielles 44 (Winter 1989): 97-119.

Kaufman, Robert S. and Roger T. Kaufman. Union Effects on Productivity, Personnel Practices, and

8/8/2019 Verma.UnionImpactOnHRM.J

verma.unionimpactonhrm.jlr2005

Documents